Characterizing the ambient seismic wavefield for upper crustal imaging: energy sources and station deployment protocols in an ocean island setting

Techniques using ambient seismic energy have become powerful tools in the imaging of the Earth's lithosphere. In this work we investigate the potential for using ambient energy to image the upper crust at the highest possible resolution. Such images are sought-after in deducing regional-scale 3D geological structure. They are critical to activities which use the Earth's crust (e.g. hot dry rock geothermal energy production, carbon sequestration) as well as placing fundamental constraints on the crustal architecture which defines tectonic structure and evolution. We have deployed a line of variously spaced seismic stations in a region of relatively well-constrained regional geology in Eastern Tasmania. Located south of mainland Australia, in the 'roaring 40's' southerly latitudes, the ambient seismic energy is of considerable amplitude. We are able to characterise the ambient seismic wavefield: investigating the influence of station deployment geometry, deployment and processing protocols, and ocean and atmospheric conditions on the amplitude and frequency content of the signals derived from correlated waveforms. We also investigate the potential of using diffuse seismic sources from a highway and railway using ambient energy techniques. We find that, in this ocean-dominated island setting, a correlated signal of sufficient strength to model for structure is obtained in a few days. The relationship between station separation and the dominant wavelengths in the correlated signals places a minimum limit on station separations which are usable with standard modelling techniques. Hence, in this environment, crustal imaging may be best carried out using deployments with overlapping, frequently moved (or numerous and short-deployed), sets of station pairs. The potential for using cultural noise sources is limited by the dominance of the natural noise sources in the ambient wavefield.